Wednesday’s IFCN Talk highlighted the work of the Duke Reporters’ Lab’s automated fact-checking project, Squash, but as director Bill Adair noted, it’s far from the only fact-checking project using automation to fight falsehoods. 

Argentine fact-checking network Chequeado launched its automated fact-checking bot Chequeabot in 2018, which uses scans of transcripts from media organizations to help identify claims for fact-checkers. A 2019 Poynter article found that roughly one of five fact checks written by Chequado started from a claim detected by its chatbot. 

British fact-checking organization Full Fact, which has collaborated with Chequado on a number of automated fact-checking projects, has been researching this technology since 2015. In addition to detection, Full Fact has used automation to keep track of how many times a false claim has been repeated, and help fact-checkers respond to claims in real time. In 2019, a collaboration between Full Fact, Chequeado and Africa Check was recognized by Google’s AI Impact challenge out of a field of 2,600 applicants. 

More recently, Spanish fact-checker Maldita.es has used automation to both speed up responses to fact-checking requests from its audience, and detect potential patterns in the types of falsehoods most commonly spread. This was especially important during the COVID-19 pandemic when requests for fact-checking skyrocketed on Maldita.es’ WhatsApp tipline. 

Aos Fatos in Brazil has had similar success using automation to both detect falsehoods and disseminate fact-checks. 

But for automation to really work, fact-checkers need help both in growing their ranks and in growing their audience.

Adair identified the need for more fact checks and more fact-checkers as a limiting factor on the effectiveness of Squash’s detection powers. In Spain, Maldita.es co-founder and CEO Clara Jiménez Cruz said her organization’s chatbot is only as powerful as its audience since it requires user submissions to pick up on patterns of falsehoods.

That’s why media literacy has been so crucial, with organizations like Chequeado, MediaWise and a host of other fact-checkers around the world putting an emphasis on training the public on the basics of fighting online falsehoods. Automation can enhance fact-checkers’ work, but only public buy-in can defeat disinformation.

If you are a fact-checker and you’d like your work/projects/achievements highlighted in the next edition, send us an email at [email protected] by next Tuesday.

Any corrections? Tips? We’d love to hear from you: [email protected]. 

Thanks for reading Factually.